The songbird has emerged as a powerful model system in which to pursue a mechanistic, neural circuits-level understanding of the development of complex sequential behaviors. Songbirds acquire their songs through a process reminiscent of speech acquisition in humans, modeling their vocalizations on the songs of their parents. They begin by singing a highly variable unstructured sequence of sounds, called subsong, which is like early babbling in humans. Gradually, their vocalizations acquire a single rhythmically repeated 'protosyllable', similar to canonical babbling in humans ('bababa'), which is then followed by a differentiation of sounds into multiple distinct syllables that will comprise their adult song. The finished product is a stereotyped sequence of sounds that is driven by a set of forebrain premotor nuclei analogous to motor cortex. The process of motor exploration, followed by the gradual emergence and refinement of precise temporally structured behavior, is the basic framework by which all complex behaviors are learned. In our previous studies, we have shown that subsong vocalizations are not simply the product of an 'immature' motor cortex, but are produced by a separate cortical circuit (a 'variability-generating circuit') dedicaed to the production of exploratory vocal variability. In preliminary experiments, we have shown that the earliest stereotyped components of juvenile song (protosyllables) are generated by rhythmic stereotyped activity in the same part of motor cortex that eventually produces adult song (the 'sequence- generating circuit'). The first specific aim of this proposal builds on earlier work to fully characterize the development of neural activity in the sequence-generating circuit and its relation to milestones in vocal learning. We have also found that the emerging stereotyped activity motor cortex is highly synchronized with the ongoing babbling vocalizations. In other words, the sculpting of syllables from babbling appears to require an intricate coordination between the variability- and sequence-generating circuits. Thus, the second two aims of this proposal are focused on characterizing signals in the thalamic and cortical pathways that link these two circuits. In summary, we have begun to understand, at a detailed mechanistic level, how sequence- generating circuits in motor cortex begin to sculpt adult-like behaviors out of the exploratory movements produced by separate variability-generating circuits. Given the homology between avian and mammalian forebrain circuitry, such an understanding could have profound implications for identifying underlying causes, at the circuit level, of human developmental disorders in both motor and cognitive domains.